Abstract
Abnormalities of red blood cell (RBC) indices may affect glycated haemoglobin (HbA1c) levels. We assessed the influence of haemoglobin (Hb) and mean corpuscular volume (MCV) on the performance of HbA1c in detecting dysglycaemia among reproductive aged women planning to conceive. Women aged 18-45 years (n = 985) were classified as normal (12 ≤ Hb ≤ 16 g/dL and 80 ≤ MCV ≤ 100 fL) and abnormal (Hb < 12 g/dL and/or MCV < 80 fL). The Area Under the Receiver Operating Characteristic (AUROC) curve was used to determine the performance of HbA1c in detecting dysglycaemic status (prediabetes and diabetes). There were 771 (78.3%) women with normal RBC indices. The AUROCs for the normal and abnormal groups were 0.75 (95% confidence interval 0.69, 0.81) and 0.80 (0.70, 0.90), respectively, and were not statistically different from one another [difference 0.04 (- 0.16, 0.08)]. Further stratification by ethnicity showed no difference between the two groups among Chinese and Indian women. However, Malay women with normal RBC indices displayed lower AUROC compared to those with abnormal RBC indices (0.71 (0.55, 0.87) vs. 0.98 (0.93, 1.00), p = 0.002). The results suggest that the performance of HbA1c in detecting dysglycaemia was not influenced by abnormal RBC indices based on low Hb and/or low MCV. However, there may be ethnic variations among them.